You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
[UX Design] Observability visualization (advanced features WIP)
Anatomy- OpenSearch viz vs other charting engines
Config panel
Visualization workspace
Chart types and feature to include:
Config panel: see table below
Visualization suggestion: include recommended chart types for selected data, and show preview (smart suggestions)
Chart type search
Prioritized charts (definition and use case)
*see below table for more chart types and configurations
Line
A line graph utilizes points and lines to represent change over time. It is a chart that shows a line joining several points or a line that shows the relation between the points. The graph represents quantitative data between two changing variables with a line or curve that joins a series of successive data points.
Bar
Bar chart uses bars to show comparisons between categories of data. The bars can be displayed vertically or horizontally depending on the use case. A bar chart always has 2 axis- one usually has numerical values, and the other one is usually used for types of categories being compared. Horizontal bar: x axis = category. Vertical bar: y axis = category.
Pie
Pie chart divides the circular statistical graphic into sectors or sections to illustrate the numerical problems. Each sector denotes a proportionate part of the whole. To find out the composition of something, pie chart works the best at that time. In some cases, pie charts can replace other graphs like the bar graph, line plots, histograms, etc.
Heat map
A heat map shows magnitude of a phenomenon as color in two dimensions. The color representation used in a heat map are hue or intensity. There are two types of heat maps- spacial and cluster. A spacial heat map is often layered on top of a physical map.
Tree map
Displays hierarchical data as a set of nested rectangles (tree structured). With the color correlation, users can see a pattern in a selected dataset.
Gauge
Displays a single data value with a quantitative context. With a shape that resembles a speedometer, gauge chart aims to track the progress of a KPI in comparison to a set target or to other time periods. It uses various colors to divide a scale into segments that represent different values of the same metric, usually min, med, max range. The needle/indicator to point at a certain value over the pivot point.
Log view
Displays log lines from data sources that support logs. Quick view from event analytics logs. Typically you would use this panel next to a graph panel to display the log output of a related process.
Table
Displays a quick view from notebooks SQL / PPL table.
Dashboard list
A dashboard list contains one or multiple links of destinations. Destinations can be another dashboard, URL, etc.
Same widget on the Event analytics home page that can be pre-filtered with type and search.
Stat
Shows the value and/or spark line.
Histogram
Displays numeric data in ranges, where each bar represents how frequently numbers fall into a particular range.
The Y-axis represents the count of values that fall into each bracket. The X-axis represents the value range.
Config Panel
Structure
Panel options- title, description
Tooltip- show, hide, position
Legend- show, hide, position
Graph styles- line/bar/point, width, opacity, gradient, connect null value
Axis- position, label, width, min/max, grid
Standard options- unit, value min/max, decimals, etc.
Data links-
Value mappings
Thresholds- color, thickness, opacity
Override
Other
Table view toggle (persistent in toolbar)
Time display per panel
Example 1 - Time Series
Line syle
Thickness
Color
Fill opacity
Edit series color directly in the legend
Line interpolation
Linear
Smooth
Step before
Step after
Axis min & max
Auto
Hard
Soft
Multiple Y axis
Add additional Y axis
Dash and dot line style settings
Adjust dash/dot space
Value mapping?
Show data points on line
Turn off line, display points only
Stacked line chart
Mixed style
line/bar/point
Connect null values?
Threshold
Multiple thresholds
Absolute, percentage
Example 2 - Bar
Bar color style
Transparent gradient
Hue gradient
Bar/line/point interchangeable in graph style
Histogram?
Nice to have (other):
Collapse all sections in config panel.
Should we include overrides in the dedicated sections?
Drilldown- dashboard level and visualization panel level
Dashboard layout- viz panel snaps when repositioning
Orientation, Rotate bar labels, Label max length, Show values, Stacking (auto, normal, 100%), Group width, Bar width, Bar radius, Line width, Fill opacity, Gradient mode (none, opacity, hue, scheme)
Chart type (pie, donut)Labels (percent, name, value)*Donut - band thickness (Plotly)*Multiple donut (Plotly)?
Show headerShow footerEnable paginationCol widthMin col widthAlignment (auto, left, center, right)Cell display mode*Cell value inspect (on/off)Col filter (on/off)
Axis
Display nameValue format Series colorPositionShow/hide line, labelWidthSoft minSoft maxShow grid lines (auto, on, off)Scale (linear, logarithmic)
Display nameValue format Series colorPositionShow/hide line, labelWidthSoft minSoft maxShow grid lines (auto, on, off)Scale (linear, logarithmic)
N/A
N/A
N/A
N/A
N/A
N/A
Y Axis:1. Unit2. Scale3. Min/Max4. Width5. DecimalsBucketsY Axis buckets, SizeX Axis buckets, SizeData format
[UX Design] Observability visualization (advanced features WIP)
Anatomy- OpenSearch viz vs other charting engines
Chart types and feature to include:
Prioritized charts (definition and use case)
*see below table for more chart types and configurations
Line
A line graph utilizes points and lines to represent change over time. It is a chart that shows a line joining several points or a line that shows the relation between the points. The graph represents quantitative data between two changing variables with a line or curve that joins a series of successive data points.
Bar
Bar chart uses bars to show comparisons between categories of data. The bars can be displayed vertically or horizontally depending on the use case. A bar chart always has 2 axis- one usually has numerical values, and the other one is usually used for types of categories being compared. Horizontal bar: x axis = category. Vertical bar: y axis = category.
Pie
Pie chart divides the circular statistical graphic into sectors or sections to illustrate the numerical problems. Each sector denotes a proportionate part of the whole. To find out the composition of something, pie chart works the best at that time. In some cases, pie charts can replace other graphs like the bar graph, line plots, histograms, etc.
Heat map
A heat map shows magnitude of a phenomenon as color in two dimensions. The color representation used in a heat map are hue or intensity. There are two types of heat maps- spacial and cluster. A spacial heat map is often layered on top of a physical map.
Tree map
Displays hierarchical data as a set of nested rectangles (tree structured). With the color correlation, users can see a pattern in a selected dataset.
Gauge
Displays a single data value with a quantitative context. With a shape that resembles a speedometer, gauge chart aims to track the progress of a KPI in comparison to a set target or to other time periods. It uses various colors to divide a scale into segments that represent different values of the same metric, usually min, med, max range. The needle/indicator to point at a certain value over the pivot point.
Log view
Displays log lines from data sources that support logs. Quick view from event analytics logs. Typically you would use this panel next to a graph panel to display the log output of a related process.
Table
Displays a quick view from notebooks SQL / PPL table.
Dashboard list
A dashboard list contains one or multiple links of destinations. Destinations can be another dashboard, URL, etc.
Same widget on the Event analytics home page that can be pre-filtered with type and search.
Stat
Shows the value and/or spark line.
Histogram
Displays numeric data in ranges, where each bar represents how frequently numbers fall into a particular range.
The Y-axis represents the count of values that fall into each bracket. The X-axis represents the value range.
Config Panel
Structure
Other
Example 1 - Time Series
Example 2 - Bar
Nice to have (other):
[Chart type & config]
Design example (Observability)
Design example (config panel section interaction)
Design example (config section features)
Related issues:
opensearch-project/observability#664
#123
opensearch-project/observability#697
opensearch-project/observability#643
#125
The text was updated successfully, but these errors were encountered: